I attended the Nova scheduler meeting yesterday morning in hopes to set the stage for a discussion about Neutron's routed networks work and how that affects Nova [1]. John suggested that I write up a backlog spec to start the discussion (as he has suggested to me before). I did so today and posted it [2]. It is mostly just a problem description for now. I plan to attend the Nova mid-cycle this month to discuss it further. However, early feedback is welcome. Carl [1] http://eavesdrop.openstack.org/meetings/nova_scheduler/2016/nova_scheduler.2016-01-04-14.00.log.html#l-60 [2] https://review.openstack.org/#/c/263898/